INDEX

Explanations

phrases related to direction or movement

oai_token-act-pair · gpt-3.5-turbo

multiple instances of the word "for" and phrases indicating direction or destination

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

FORE

-0.71

 persisted

-0.70

nces

-0.68

 belong

-0.65

Used

-0.65

inant

-0.64

 applied

-0.64

 complied

-0.64

Rated

-0.62

 secondly

-0.62

POSITIVE LOGITS

 Charlottesville

0.80

pex

0.77

 retirement

0.77

ä»£

0.76

 Skydragon

0.74

 Showdown

0.71

 town

0.69

 Downing

0.68

tnc

0.67

 Munich

0.67

Activations Density 0.066%

phrases related to direction or movement

multiple instances of the word "for" and phrases indicating direction or destination

No Comments

No Known Activations